Baseline - LSTM-GAN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "a80a379b-932c-4401-8d76-b5ae5dedc6cc",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import random\n",
    "\n",
    "os.environ[\"CUDA_DEVICE_ORDER\"] = \"PCI_BUS_ID\";\n",
    "os.environ[\"CUDA_VISIBLE_DEVICES\"] = \"1\"\n",
    "os.environ[\"TF_FORCE_GPU_ALLOW_GROWTH\"] = \"true\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "saving-behavior",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import tensorflow as tf\n",
    "\n",
    "from evaluator import evaluate\n",
    "from data_loader import load_kdd_cup_urc, load_yahoo_A1, load_yahoo_A2, load_yahoo_A3, load_yahoo_A4, load_power_demand # Univariate Datasets\n",
    "from data_loader import load_nasa, load_ecg, load_gesture, load_smd # Multivariate Datasets\n",
    "\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras import layers\n",
    "from tqdm.notebook import tqdm\n",
    "\n",
    "# THESE LINES ARE FOR REPRODUCIBILITY\n",
    "random.seed(0)\n",
    "np.random.seed(0)\n",
    "tf.random.set_seed(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "49c0a8e3-705d-429b-b7a0-c3c26a3a7b2b",
   "metadata": {},
   "outputs": [],
   "source": [
    "def train_gan(gan, dataset, batch_size, codings_size, dim, n_epochs=50):\n",
    "    generator, discriminator = gan.layers\n",
    "    for epoch in range(n_epochs):\n",
    "        for X_batch in dataset:\n",
    "            X_batch = tf.cast(X_batch, tf.float32)\n",
    "            # phase 1 - training the discriminator\n",
    "            noise = tf.random.normal(shape=[batch_size, codings_size, dim])\n",
    "            generated_images = generator(noise)\n",
    "            X_fake_and_real = tf.concat([generated_images, X_batch], axis=0)\n",
    "            y1 = tf.constant([[0.]] * batch_size + [[1.]] * batch_size)\n",
    "            discriminator.trainable = True\n",
    "            discriminator.train_on_batch(X_fake_and_real, y1)\n",
    "            # phase 2 - training the generator\n",
    "            noise = tf.random.normal(shape=[batch_size, codings_size, dim])\n",
    "            y2 = tf.constant([[1.]] * batch_size)\n",
    "            discriminator.trainable = False\n",
    "            gan.train_on_batch(noise, y2)\n",
    "            \n",
    "def get_gan(x_train):\n",
    "    generator = keras.models.Sequential([\n",
    "        layers.InputLayer(input_shape=(x_train.shape[1], x_train.shape[2])),\n",
    "        layers.LSTM(128, return_sequences=True),\n",
    "        layers.Dropout(rate=0.2),\n",
    "        layers.LSTM(64),\n",
    "        layers.RepeatVector(x_train.shape[1]),\n",
    "        layers.LSTM(64, return_sequences=True),\n",
    "        layers.Dropout(rate=0.2),\n",
    "        layers.LSTM(128),\n",
    "        layers.Dense(x_train.shape[1] * x_train.shape[2]),\n",
    "        layers.Reshape([x_train.shape[1], x_train.shape[2]])\n",
    "    ])\n",
    "\n",
    "    discriminator = keras.models.Sequential([\n",
    "        layers.InputLayer(input_shape=(x_train.shape[1], x_train.shape[2])),\n",
    "        keras.layers.LSTM(64, return_sequences=True),\n",
    "        keras.layers.Dropout(0.2),\n",
    "        keras.layers.LSTM(64),\n",
    "        keras.layers.Dropout(0.2),\n",
    "        keras.layers.Dense(1, activation=\"sigmoid\")\n",
    "    ])\n",
    "\n",
    "    gan = keras.models.Sequential([generator, discriminator])\n",
    "\n",
    "    discriminator.compile(loss=\"binary_crossentropy\", optimizer=\"adam\")\n",
    "    discriminator.trainable = False\n",
    "    gan.compile(loss=\"binary_crossentropy\", optimizer=\"adam\")\n",
    "    return gan"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fa97c067-d18d-4c06-ae1b-2cbfe88788b6",
   "metadata": {},
   "source": [
    "### Yahoo S5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "eight-partition",
   "metadata": {},
   "outputs": [],
   "source": [
    "total_scores = {'dataset': [], 'f1': [], 'pr_auc': [], 'roc_auc': []}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "71866d92-1163-4975-bb99-4302ccda66c3",
   "metadata": {
    "scrolled": true,
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8cc131fe3d06473494a6302ca0f300af",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/67 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "98d5775ff4d344dfb7f18277a2dd1555",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/56 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "yahoo_A1 0.03738317383177589 0.00480769253887533 -0.0\n",
      "yahoo_A1 0.05504586610554703 0.009569596976538494 -0.0\n",
      "yahoo_A1 0.6666666074074105 0.35161276324797885 0.49999999116666693\n",
      "yahoo_A1 0.018691586898419075 0.0047169806559274234 -0.0\n",
      "yahoo_A1 0.03703703333333351 0.004761904990972433 -0.0\n",
      "yahoo_A1 0.03703703333333351 0.004761904990972433 -0.0\n",
      "yahoo_A1 0.7999999200000041 0.34487175816617643 0.8206348848004553\n",
      "yahoo_A1 0.8928570899234721 0.8925086837075218 0.9321915662807525\n",
      "yahoo_A1 0.8965516711058291 0.8991838931229897 0.9464302088610026\n",
      "yahoo_A1 0.7499999343750033 0.43974901995979976 0.8913366053109713\n",
      "yahoo_A1 0.10810809768687688 0.00995049562920052 -0.0\n",
      "yahoo_A1 0.09090908206611638 0.004950496190208909 -0.0\n",
      "yahoo_A1 0.9677418792924064 0.8855291095785212 0.9469803300327997\n",
      "yahoo_A1 0.7999999200000041 0.3452532625488612 0.8235293756632082\n",
      "yahoo_A1 0.7499999343750033 0.41403121761016803 0.678883478370421\n",
      "yahoo_A1 0.9523808934240391 0.8297355505257258 0.9339712821434493\n",
      "yahoo_A1 0.9499999453750025 0.6966953920866147 0.9536274450609777\n",
      "yahoo_A1 0.07339448820806381 0.009756329900136602 -0.0\n",
      "yahoo_A1 0.6666666000000033 0.2755851085420757 0.7357025928823455\n",
      "yahoo_A1 0.33333329444444565 0.013040019966263578 0.5100659673285961\n",
      "yahoo_A1 0.054054048697346334 0.009389878237270773 -0.0\n",
      "yahoo_A1 0.03636363272727291 0.009302526352111108 -0.0\n",
      "yahoo_A1 0.749999943750003 0.6070753023179527 0.8495029130472014\n",
      "yahoo_A1 0.2644627865309766 0.07920330115282279 0.24499423579061314\n",
      "yahoo_A1 0.018691586898419075 0.0047169806559274234 -0.0\n",
      "yahoo_A1 0.12499998805803669 0.005050506538735456 -0.0\n",
      "yahoo_A1 0.1880341706771875 0.045987433198940664 0.06236842031176133\n",
      "yahoo_A1 0.1333332871111258 0.02505745323587545 0.4282828022475276\n",
      "yahoo_A1 0.097560965853659 0.014285715455172668 0.6633663015709264\n",
      "yahoo_A1 0.999999930000003 0.7999999503333362 0.9999999790000004\n",
      "yahoo_A1 0.9333332794074098 0.5761429494542514 0.9101949395384419\n",
      "yahoo_A1 0.39999995200000177 0.012757476583328748 0.5575979956032575\n",
      "yahoo_A1 0.9999999428571453 0.9285714024533185 0.9999999917701865\n",
      "yahoo_A1 0.14035086389658463 0.04071568712401419 0.32482992627608887\n",
      "yahoo_A1 0.9999999000000052 0.4999999375000063 0.9999999490384642\n",
      "yahoo_A1 0.07272726558677738 0.00847457695282729 0.4411764346164968\n",
      "yahoo_A1 0.10810809768687688 0.0076535934550981295 0.0227272708103766\n",
      "yahoo_A1 0.018691586898419075 0.0047169806559274234 -0.0\n",
      "yahoo_A1 0.09090908206611638 0.004950496190208909 -0.0\n",
      "yahoo_A1 0.33333329444444565 0.011497435157748941 0.37186466909565197\n",
      "yahoo_A1 0.7499999343750033 0.4178043313293256 0.7291261944587777\n",
      "yahoo_A1 0.1333332835555728 0.022271571922094315 0.3920067774684759\n",
      "yahoo_A1 0.13333328177779635 0.027085055463994413 0.2737487769940665\n",
      "yahoo_A1 0.07272726558677738 0.004854369994174695 -0.0\n",
      "yahoo_A1 0.9523808934240391 0.8699494645255301 0.9703349177985168\n",
      "yahoo_A1 0.03703703333333351 0.004761904990972433 -0.0\n",
      "yahoo_A1 0.7499999343750033 0.5580356729834206 0.9820388016347532\n",
      "yahoo_A1 0.0 0.0 -0.0\n",
      "yahoo_A1 0.10526314755309392 0.009615385836948672 -0.0\n",
      "yahoo_A1 0.03703703333333351 0.004761904990972433 -0.0\n",
      "yahoo_A1 0.19999997800000044 0.030783737089413146 0.16427951147694958\n",
      "yahoo_A1 0.9166666093750028 0.8994082563045758 0.99665550883238\n",
      "yahoo_A1 0.9473683612188393 0.8113051160057393 0.9231578842762083\n",
      "yahoo_A1 0.22033896307095824 0.0534650904444822 0.049039305546747805\n",
      "yahoo_A1 0.9999999250000032 0.7499999447916703 0.9999999740099017\n",
      "yahoo_A1 0.7999999200000041 0.34095458430995584 0.7285714004965999\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "32cad5a966f44505870ac3ede6e8e154",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "01e5a4944fee40b09bd51e0a445c3789",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.005194840631162036 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.0026041674459758115 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.005869573522519201 0.06806282426628486\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.006442907630584995 0.24978183750193275\n",
      "yahoo_A2 0.06829267626412905 0.005194840631162036 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.00549450571758334 0.5135134618845925\n",
      "yahoo_A2 0.06829267626412905 0.007352942187467447 0.6492146322909886\n",
      "yahoo_A2 0.029850743239028747 0.007575756806703475 0.4961239810107616\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.005485113979077226 0.0929319324775228\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.0039267021522734705 0.005235601698829523\n",
      "yahoo_A2 0.020408161203665145 0.005154638654479805 0.25581392770867384\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.002645503420634792 0.01570680470381858\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.0026041674459758115 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.005077285413257899 0.24214657860358968\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.004950495328120273 0.4594594132651617\n",
      "yahoo_A2 0.041450773099949295 0.0027173918096981384 0.01081080972388618\n",
      "yahoo_A2 0.06829267626412905 0.0026041674459758115 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.005194840656018777 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.0026041674459758115 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0030120486683180674 0.10810809723886158\n",
      "yahoo_A2 0.06829267626412905 0.002604167445975811 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.002777778557145538 0.06544502182437463\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.0026041674459758115 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.0026041674459758115 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.005194840631162036 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.005194840631162036 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0030674851889766104 0.12972972290284918\n",
      "yahoo_A2 0.06829267626412905 0.002604167445975811 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.00692733412829712 0.2657067928582694\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.005194840631162036 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.004032258370484449 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.0026041674459758115 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.002604167445975811 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.39999995200000177 0.014166285468757837 0.5882882470881658\n",
      "yahoo_A2 0.06829267626412905 0.015446400380925287 0.5235601916979621\n",
      "yahoo_A2 0.06829267626412905 0.0031055908032185985 0.1623036501768057\n",
      "yahoo_A2 0.9999998500000123 0.0 0.9999998992248162\n",
      "yahoo_A2 0.041450773099949295 0.002702703209537798 0.00540540486194309\n",
      "yahoo_A2 0.06829267626412905 0.0026041674459758115 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.0039062505833184985 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.0026041674500134598 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.002673797569785158 0.028795808885447485\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.041450773099949295 0.0026881725513145803 -0.0\n",
      "yahoo_A2 0.06829267626412905 0.0026041674459758115 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n",
      "yahoo_A2 0.39999995200000177 0.008748454258354394 0.4445945734646346\n",
      "yahoo_A2 0.06829267626412905 0.0026041674459758115 -0.0\n",
      "yahoo_A2 0.015267174034147195 0.0038461534585799203 -0.0\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9978cf20450d4a2eac205636190113f7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "db8a640c75ec47bd8d7345a64c926cf0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/92 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "yahoo_A3 0.03174602857142873 0.004065040847075319 -0.0\n",
      "yahoo_A3 0.2499999562500052 0.0645205608785468 0.6574712477172786\n",
      "yahoo_A3 0.4999999500000025 0.24689536094108136 0.8172316131487688\n",
      "yahoo_A3 0.03174602857142873 0.004065040847075319 -0.0\n",
      "yahoo_A3 0.14925371730897866 0.018608312721908216 0.37646197586372765\n",
      "yahoo_A3 0.0624999938476567 0.0041322321806666834 -0.0\n",
      "yahoo_A3 0.23529406643599615 0.13527981687896926 0.7653737514739931\n",
      "yahoo_A3 0.28571425306122533 0.015975360951210008 0.49406775916675394\n",
      "yahoo_A3 0.28571425306122533 0.06582845084432447 0.6706214419997101\n",
      "yahoo_A3 0.18181816198347142 0.011662114977797323 0.2684140791165204\n",
      "yahoo_A3 0.14925371730897866 0.004347827665800764 -0.0\n",
      "yahoo_A3 0.14925371730897866 0.02852864082976612 0.26701649476294287\n",
      "yahoo_A3 0.03174602857142873 0.004065040847075319 -0.0\n",
      "yahoo_A3 0.17647057188581453 0.006987713390723464 0.05580356613321154\n",
      "yahoo_A3 0.04724408980097991 0.004098361191078604 -0.0\n",
      "yahoo_A3 0.18181816198347142 0.023996986780381327 0.5637461221121852\n",
      "yahoo_A3 0.06451611654526714 0.013250318731787761 0.44672128262564037\n",
      "yahoo_A3 0.03174602857142873 0.004065040847075319 -0.0\n",
      "yahoo_A3 0.0624999938476567 0.007901892195965213 0.2249999785625021\n",
      "yahoo_A3 0.03174602857142873 0.004347826254870989 0.06557376388067784\n",
      "yahoo_A3 0.03174602857142873 0.004065040847075319 -0.0\n",
      "yahoo_A3 0.0624999938476567 0.0041322321806666834 -0.0\n",
      "yahoo_A3 0.09230768336094745 0.004201681792727941 -0.0\n",
      "yahoo_A3 0.3076922556213091 0.09456570656697966 0.8443502425104398\n",
      "yahoo_A3 0.03174602857142873 0.004237288314751278 0.045081962667965984\n",
      "yahoo_A3 0.09230768336094745 0.004201681792727941 -0.0\n",
      "yahoo_A3 0.0624999938476567 0.0041322321806666834 -0.0\n",
      "yahoo_A3 0.03174602857142873 0.004065040847075319 -0.0\n",
      "yahoo_A3 0.06896549298455006 0.01350865642301344 0.3805555314247704\n",
      "yahoo_A3 0.2222221975308647 0.03975729093358548 0.5480090173759447\n",
      "yahoo_A3 0.20289853219911938 0.014141834697453042 0.09999999730303041\n",
      "yahoo_A3 0.14925371730897866 0.014615570246745731 0.17799706932181078\n",
      "yahoo_A3 0.153846101775164 0.03276705939058109 0.6224137412656442\n",
      "yahoo_A3 0.3636363272727291 0.2018727720163585 0.7441511248840607\n",
      "yahoo_A3 0.12499995156251753 0.016224284914768898 0.22923727993967696\n",
      "yahoo_A3 0.14925371730897866 0.0046728987420123404 0.07017543151739064\n",
      "yahoo_A3 0.09230768336094745 0.012194422742404032 0.43926549564940237\n",
      "yahoo_A3 0.12121210964187425 0.004273505647355304 -0.0\n",
      "yahoo_A3 0.0624999938476567 0.004201681440916229 0.016666664986111218\n",
      "yahoo_A3 0.17647057188581453 0.016970439503421857 0.26605451055112095\n",
      "yahoo_A3 0.06666666000000032 0.01315789469492407 0.7008196287288395\n",
      "yahoo_A3 0.12121210964187425 0.004310346197689935 0.008620688785671996\n",
      "yahoo_A3 0.03174602857142873 0.004065040847075319 -0.0\n",
      "yahoo_A3 0.09230768336094745 0.01820299816549018 0.4485875506033637\n",
      "yahoo_A3 0.15384610414202635 0.02209630157929952 0.23436986077005625\n",
      "yahoo_A3 0.17647057188581453 0.006676676511255319 0.006696428007413917\n",
      "yahoo_A3 0.1897810044434987 0.02007180823935356 0.1861998313455226\n",
      "yahoo_A3 0.12121210964187425 0.012637570596403625 0.30603447121606275\n",
      "yahoo_A3 0.666666577777783 0.019999997526478106 0.9016392582773531\n",
      "yahoo_A3 0.0624999938476567 0.005747127047809303 0.28333330476389174\n",
      "yahoo_A3 0.0624999938476567 0.009019468370862068 0.24791665086979298\n",
      "yahoo_A3 0.09230768336094745 0.015355443076219836 0.2138418019567256\n",
      "yahoo_A3 0.09230768336094745 0.007717116088922289 0.09604518993743885\n",
      "yahoo_A3 0.12121210964187425 0.019526696068476783 0.4347700993469421\n",
      "yahoo_A3 0.12121210964187425 0.006410257565866539 0.34482755142687627\n",
      "yahoo_A3 0.399999960000002 0.14895124214252245 0.39246715070046695\n",
      "yahoo_A3 0.4999999500000025 0.18103251457252198 0.46610168131874674\n",
      "yahoo_A3 0.03174602857142873 0.0049019608957896026 0.1762294904293218\n",
      "yahoo_A3 0.3999999460000063 0.16064156870200244 0.5947925271873177\n",
      "yahoo_A3 0.12121210964187425 0.006505648177222858 0.021551723041765815\n",
      "yahoo_A3 0.6666666000000033 0.32881390398172017 0.9236110853182877\n",
      "yahoo_A3 0.28571425306122533 0.028298687494515095 0.5073446159980137\n",
      "yahoo_A3 0.03448275184304499 0.010019840805626732 0.40983603122144896\n",
      "yahoo_A3 0.14925371730897866 0.026199410667987248 0.2683740542633889\n",
      "yahoo_A3 0.15384611597633918 0.04579423299989593 0.4489417877165913\n",
      "yahoo_A3 0.17647057188581453 0.0044247805153377435 -0.0\n",
      "yahoo_A3 0.0624999938476567 0.0048543696342351825 0.14999998487500155\n",
      "yahoo_A3 0.22857140800000167 0.022950354203982334 0.18211755694210618\n",
      "yahoo_A3 0.13533833304313522 0.02031004739944512 0.2989854898213661\n",
      "yahoo_A3 0.03174602857142873 0.004065040847075319 -0.0\n",
      "yahoo_A3 0.39999995200000177 0.012353581414203825 0.564583292446184\n",
      "yahoo_A3 0.11111106543211643 0.027203895476455985 0.49251410398963125\n",
      "yahoo_A3 0.18181816198347142 0.025121588022767538 0.42143203934552287\n",
      "yahoo_A3 0.12121210964187425 0.004310346197689935 0.008620688785671996\n",
      "yahoo_A3 0.06666665377777961 0.023015870991241187 0.7479507458008672\n",
      "yahoo_A3 0.03174602857142873 0.004065040847075319 -0.0\n",
      "yahoo_A3 0.0624999938476567 0.004950495842027627 0.17083332319097289\n",
      "yahoo_A3 0.6666666074074105 0.4297878572306539 0.8226694674557188\n",
      "yahoo_A3 0.0624999938476567 0.005263158555227111 0.21666664481944664\n",
      "yahoo_A3 0.12121210964187425 0.004347827453247221 0.01724137757134388\n",
      "yahoo_A3 0.13533833304313522 0.004310346308505339 -0.0\n",
      "yahoo_A3 0.12121210964187425 0.01624577116504675 0.2366379248956188\n",
      "yahoo_A3 0.12121210964187425 0.004273505647355304 -0.0\n",
      "yahoo_A3 0.39999995200000177 0.007358064817934129 0.3329860970170726\n",
      "yahoo_A3 0.14285711734694262 0.029707475437878373 0.5652777557210658\n",
      "yahoo_A3 0.10687021872851314 0.004237289390589688 -0.0\n",
      "yahoo_A3 0.0624999938476567 0.0041322321806666834 -0.0\n",
      "yahoo_A3 0.14925371730897866 0.034861402104923875 0.3123294287594592\n",
      "yahoo_A3 0.09230768336094745 0.01984730058796835 0.6260592723066631\n",
      "yahoo_A3 0.39999995200000177 0.009343483341414676 0.3423610981637737\n",
      "yahoo_A3 0.09230768336094745 0.019896255270014162 0.5190677659089842\n",
      "yahoo_A3 0.4999999472222262 0.20053525309669198 0.7292179632781641\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "40f80d08a9574bdaa6269fe23ff201d4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "93dc26b051a046878fc220276bcf5ec8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/82 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "yahoo_A4 0.28571425306122533 0.012039178025323623 0.28015535811604103\n",
      "yahoo_A4 0.03174602857142873 0.004065040847075319 -0.0\n",
      "yahoo_A4 0.0624999938476567 0.004237288900619699 0.024999997479166938\n",
      "yahoo_A4 0.0624999938476567 0.012001883167080892 0.33333331638888986\n",
      "yahoo_A4 0.12121210964187425 0.024817306459083973 0.34393471942529463\n",
      "yahoo_A4 0.12121210964187425 0.009883276303566947 0.18031608568387536\n",
      "yahoo_A4 0.0624999938476567 0.0041322321806666834 -0.0\n",
      "yahoo_A4 0.07751937227330145 0.008368347676241844 0.008403360917308056\n",
      "yahoo_A4 0.09230768336094745 0.013796373156142065 0.2252118564224903\n",
      "yahoo_A4 0.0624999938476567 0.0041322321806666834 -0.0\n",
      "yahoo_A4 0.12121210964187425 0.014731034933751561 0.1985016352888609\n",
      "yahoo_A4 0.18181813223141666 0.032419894454120385 0.6232638371686969\n",
      "yahoo_A4 0.39999995200000177 0.011504515546446992 0.5680555181261604\n",
      "yahoo_A4 0.09230768336094745 0.004901961833469736 0.1440677820813\n",
      "yahoo_A4 0.03174602857142873 0.004065040847075319 -0.0\n",
      "yahoo_A4 0.666666577777783 0.013701199714947385 0.854508117844673\n",
      "yahoo_A4 0.13533833304313522 0.020060200832246647 0.11804347556365787\n",
      "yahoo_A4 0.12121210964187425 0.007326519073392576 0.14008619524561647\n",
      "yahoo_A4 0.14925371730897866 0.005000001529673948 0.13157893541089677\n",
      "yahoo_A4 0.09230768336094745 0.004201681792727941 -0.0\n",
      "yahoo_A4 0.03174602857142873 0.004065040847075319 -0.0\n",
      "yahoo_A4 0.0624999938476567 0.00855275471593172 0.21249998496180678\n",
      "yahoo_A4 0.1999999480000115 0.04000419544020968 0.34519773174182894\n",
      "yahoo_A4 0.2499999562500052 0.045798895282020025 0.4877048928379482\n",
      "yahoo_A4 0.0624999938476567 0.004132232212911846 -0.0\n",
      "yahoo_A4 0.28571425306122533 0.06211206387939994 0.657485859921858\n",
      "yahoo_A4 0.35294114325259707 0.18687229290216686 0.4320125260475146\n",
      "yahoo_A4 0.03174602857142873 0.004065040847075319 -0.0\n",
      "yahoo_A4 0.14925371730897866 0.017780696972444828 0.1775010396287567\n",
      "yahoo_A4 0.18181816198347142 0.017866293091463162 0.1685063320375114\n",
      "yahoo_A4 0.6666666000000033 0.26513602891689014 0.6805555302893529\n",
      "yahoo_A4 0.09230768336094745 0.004424779858826265 0.050847452499282395\n",
      "yahoo_A4 0.03174602857142873 0.004065040847075319 -0.0\n",
      "yahoo_A4 0.39999995200000177 0.0724585639448473 0.6267360889945033\n",
      "yahoo_A4 0.24999994765625974 0.06795014578988685 0.5084817732965629\n",
      "yahoo_A4 0.399999960000002 0.18001681941019948 0.6684523535696437\n",
      "yahoo_A4 0.09230768336094745 0.010828030918494954 0.3622881062254647\n",
      "yahoo_A4 0.28571425306122533 0.009264992330280965 0.3382768176205767\n",
      "yahoo_A4 0.09230768336094745 0.004201681792727941 -0.0\n",
      "yahoo_A4 0.09523806984127565 0.019689761678276425 0.46249997218402966\n",
      "yahoo_A4 0.09230768336094745 0.004237289252431411 0.008474575416547066\n",
      "yahoo_A4 0.015999998387200003 0.004032257658038542 -0.0\n",
      "yahoo_A4 0.24999997187500062 0.02549621100666641 0.4666056024136572\n",
      "yahoo_A4 0.0624999938476567 0.0041322321806666834 -0.0\n",
      "yahoo_A4 0.0624999938476567 0.004237288900619699 0.024999997479166938\n",
      "yahoo_A4 0.03174602857142873 0.004065040847075319 -0.0\n",
      "yahoo_A4 0.08695649224953653 0.026414188586531884 0.42986109880266243\n",
      "yahoo_A4 0.03174602857142873 0.004065040847075319 -0.0\n",
      "yahoo_A4 0.07751937227330145 0.004587156884661844 0.0966386457171113\n",
      "yahoo_A4 0.22222216790124563 0.032726531889197254 0.3583333171493065\n",
      "yahoo_A4 0.16296294775308767 0.020154077056397116 0.2781921537670317\n",
      "yahoo_A4 0.12121210964187425 0.007918418774899705 0.0696839051063551\n",
      "yahoo_A4 0.03174602857142873 0.004065040847075319 -0.0\n",
      "yahoo_A4 0.666666577777783 0.013335702695725652 0.8504097578070485\n",
      "yahoo_A4 0.09230768336094745 0.004201681792727941 -0.0\n",
      "yahoo_A4 0.0624999938476567 0.004587156692724974 0.09999998991666764\n",
      "yahoo_A4 0.0624999938476567 0.0041322321806666834 -0.0\n",
      "yahoo_A4 0.0624999938476567 0.0041322321806666834 -0.0\n",
      "yahoo_A4 0.03174602857142873 0.004098360849020119 0.008196720485084744\n",
      "yahoo_A4 0.12121210964187425 0.004273505647355304 -0.0\n",
      "yahoo_A4 0.0624999938476567 0.0042372889006197 0.024999997479166938\n",
      "yahoo_A4 0.12499994921876939 0.020938021617687704 0.4674671289831559\n",
      "yahoo_A4 0.12121210964187425 0.014809119349573172 0.38146549568408294\n",
      "yahoo_A4 0.1904761387755233 0.038812174774321945 0.38517473339652464\n",
      "yahoo_A4 0.03174602857142873 0.004065040847075319 -0.0\n",
      "yahoo_A4 0.09230768336094745 0.015550646954406222 0.1889830449796756\n",
      "yahoo_A4 0.39999995200000177 0.007576052608769328 0.3868055325480343\n",
      "yahoo_A4 0.03174602857142873 0.0045871561068565525 0.1147540867911863\n",
      "yahoo_A4 0.17647057188581453 0.018528290119854393 0.03660714195838648\n",
      "yahoo_A4 0.12121210964187425 0.013126025080814228 0.014367815456962682\n",
      "yahoo_A4 0.03174602857142873 0.004132231594798262 0.01639344097016949\n",
      "yahoo_A4 0.16296294775308767 0.004385966583213805 -0.0\n",
      "yahoo_A4 0.12121210964187425 0.011468631396369812 0.38793099751040794\n",
      "yahoo_A4 0.3157894216066558 0.07624530329402976 0.39860659475094085\n",
      "yahoo_A4 0.0624999938476567 0.004132232180666684 -0.0\n",
      "yahoo_A4 0.13793100071344394 0.034515685398868935 0.46087569186554844\n",
      "yahoo_A4 0.39999995200000177 0.009027518838991045 0.44479163989843956\n",
      "yahoo_A4 0.09230768336094745 0.004201681792727941 -0.0\n",
      "yahoo_A4 0.14285711734694262 0.027129046950013745 0.43749998779513927\n",
      "yahoo_A4 0.14925371730897866 0.018182992398125625 0.10818713119408542\n",
      "yahoo_A4 0.14925371730897866 0.004347827665800764 -0.0\n",
      "yahoo_A4 0.04724408980097991 0.004237288656809763 0.03305784790656404\n"
     ]
    }
   ],
   "source": [
    "for loader in [load_yahoo_A1, load_yahoo_A2, load_yahoo_A3, load_yahoo_A4]:\n",
    "    datasets = loader(8, 4)\n",
    "    x_trains, x_tests, y_tests = datasets['x_train'], datasets['x_test'], datasets['y_test']\n",
    "    \n",
    "    for i in tqdm(range(len(x_trains))):\n",
    "        tf.keras.backend.clear_session()\n",
    "\n",
    "        X_train = x_trains[i]\n",
    "        X_test = x_tests[i]\n",
    "        \n",
    "        gan = get_gan(X_train)\n",
    "        dataset = tf.data.Dataset.from_tensor_slices(X_train)\n",
    "        dataset = dataset.batch(128, drop_remainder=True).prefetch(1)\n",
    "\n",
    "        train_gan(gan, dataset, 128, X_train.shape[1], X_train.shape[2])\n",
    "        \n",
    "        X_test_rec = gan.layers[0].predict(X_test)\n",
    "        scores = evaluate(X_test, X_test_rec, y_tests[i], is_reconstructed=True)\n",
    "    \n",
    "        total_scores['dataset'].append(loader.__name__.replace('load_', ''))\n",
    "        total_scores['f1'].append(np.max(scores['f1']))\n",
    "        total_scores['pr_auc'].append(scores['pr_auc'])\n",
    "        total_scores['roc_auc'].append(scores['roc_auc'])\n",
    "        print(loader.__name__.replace('load_', ''), np.max(scores['f1']), scores['pr_auc'], scores['roc_auc'])   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "bd9e37fd-124c-4f5c-8f1f-a9d61a4b0315",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>f1</th>\n",
       "      <th>pr_auc</th>\n",
       "      <th>roc_auc</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dataset</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>yahoo_A1</th>\n",
       "      <td>0.423290</td>\n",
       "      <td>0.266816</td>\n",
       "      <td>0.442129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>yahoo_A2</th>\n",
       "      <td>0.058884</td>\n",
       "      <td>0.003835</td>\n",
       "      <td>0.063807</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>yahoo_A3</th>\n",
       "      <td>0.165641</td>\n",
       "      <td>0.035319</td>\n",
       "      <td>0.288456</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>yahoo_A4</th>\n",
       "      <td>0.152447</td>\n",
       "      <td>0.021632</td>\n",
       "      <td>0.211376</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                f1    pr_auc   roc_auc\n",
       "dataset                               \n",
       "yahoo_A1  0.423290  0.266816  0.442129\n",
       "yahoo_A2  0.058884  0.003835  0.063807\n",
       "yahoo_A3  0.165641  0.035319  0.288456\n",
       "yahoo_A4  0.152447  0.021632  0.211376"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yahoo_results = pd.DataFrame(total_scores)\n",
    "yahoo_results.groupby('dataset').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cb7c5275-3d66-4e5e-9585-dd33b88bcbf1",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "8890a678-6116-4222-b0ff-395fcd7f41aa",
   "metadata": {},
   "source": [
    "### NASA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "e4387a80-585a-488e-94a8-546f42d47632",
   "metadata": {},
   "outputs": [],
   "source": [
    "total_scores = {'dataset': [], 'f1': [], 'pr_auc': [], 'roc_auc': []}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "106a238d-c5cb-49e9-9764-b1bc37f09436",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3b27e06f92054752a54322a070020aae",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1efb1bf9be2d42aa8002dfb031c13cf4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "D1 0.2295534087061448 0.11135287422174965 0.42432860348719503\n",
      "D2 0.19498578202833744 0.18883812431052427 0.5218225683887379\n"
     ]
    }
   ],
   "source": [
    "for loader in [load_nasa]:\n",
    "    datasets = loader(8, 4)\n",
    "    x_trains, x_tests, y_tests = datasets['x_train'], datasets['x_test'], datasets['y_test']\n",
    "    \n",
    "    for i in tqdm(range(len(x_trains))):\n",
    "        tf.keras.backend.clear_session()\n",
    "\n",
    "        X_train = x_trains[i]\n",
    "        X_test = x_tests[i]\n",
    "        \n",
    "        gan = get_gan(X_train)\n",
    "        dataset = tf.data.Dataset.from_tensor_slices(X_train)\n",
    "        dataset = dataset.batch(128, drop_remainder=True).prefetch(1)\n",
    "\n",
    "        train_gan(gan, dataset, 128, X_train.shape[1], X_train.shape[2])\n",
    "        \n",
    "        X_test_rec = gan.layers[0].predict(X_test)\n",
    "        scores = evaluate(X_test, X_test_rec, y_tests[i], is_reconstructed=True)\n",
    "        \n",
    "        total_scores['dataset'].append(f'D{i+1}')\n",
    "        total_scores['f1'].append(np.max(scores['f1']))\n",
    "        total_scores['pr_auc'].append(scores['pr_auc'])\n",
    "        total_scores['roc_auc'].append(scores['roc_auc'])\n",
    "        print(f'D{i+1}', np.max(scores['f1']), scores['pr_auc'], scores['roc_auc'])   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "abb7a6a5-3924-4a33-bc54-871559819602",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>f1</th>\n",
       "      <th>pr_auc</th>\n",
       "      <th>roc_auc</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dataset</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>D1</th>\n",
       "      <td>0.229553</td>\n",
       "      <td>0.111353</td>\n",
       "      <td>0.424329</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>D2</th>\n",
       "      <td>0.194986</td>\n",
       "      <td>0.188838</td>\n",
       "      <td>0.521823</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               f1    pr_auc   roc_auc\n",
       "dataset                              \n",
       "D1       0.229553  0.111353  0.424329\n",
       "D2       0.194986  0.188838  0.521823"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nasa_results = pd.DataFrame(total_scores)\n",
    "nasa_results.groupby('dataset').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5b040420-82ec-4d83-be2c-44963aa0a427",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "41754734-5c75-4caf-9069-fb3b39c7a0da",
   "metadata": {},
   "source": [
    "### SMD"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "60a26807-8359-42d4-82b4-9e1d37b6acaa",
   "metadata": {},
   "outputs": [],
   "source": [
    "total_scores = {'dataset': [], 'f1': [], 'pr_auc': [], 'roc_auc': []}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "41af318e-1b08-4cc9-a680-6b4a978544f2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f6372649ae6e4328b5a63304c4f03094",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/28 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "86710bb1bddf43d0a6a6b3f70aed3498",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/28 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "smd 0.17627464443920743 0.5532991977145021 0.5087209301507045\n",
      "smd 0.54014593560659 0.4462999109290859 0.8824795235377808\n",
      "smd 0.12276212451515037 0.06572709066824192 0.6084192505260958\n",
      "smd 0.1720429644994407 0.12994413452312695 0.5911213737600066\n",
      "smd 0.6551723644470901 0.6494110815795878 0.9003509342585144\n",
      "smd 0.8189157523251666 0.8089249417333806 0.8878390602911748\n",
      "smd 0.19741786078160556 0.12100973496617938 0.497259602167716\n",
      "smd 0.26562497675781443 0.2130484213122197 0.72465250898784\n",
      "smd 0.3119657678236237 0.2531483104973473 0.5859147942418421\n",
      "smd 0.21837868410447125 0.2010120365146047 0.5086331327245738\n",
      "smd 0.5439999556096035 0.6077741926201783 0.699546425000602\n",
      "smd 0.31284913550347165 0.30551668132831866 0.7502333043621344\n",
      "smd 0.2362637000452899 0.13904086802822196 0.2502272623199936\n",
      "smd 0.4545454181312221 0.3717485758955818 0.680554448698186\n",
      "smd 0.6812226593085595 0.5800075104817434 0.8392719046972075\n",
      "smd 0.9499999477500025 0.8942184364352316 0.9760487986827799\n",
      "smd 0.5199361646034791 0.6694698314706001 0.677494251386233\n",
      "smd 0.49557518220691027 0.6335570090835618 0.6666131701364144\n",
      "smd 0.817995861126378 0.7201526114727894 0.8292248747764801\n",
      "smd 0.4473683798822749 0.34593053353488257 0.8768172553124861\n",
      "smd 0.16254921148336762 0.07103510637354454 0.6494505656891384\n",
      "smd 0.06847081177440445 0.02107579074441389 0.3374862396326539\n",
      "smd 0.8431771392171125 0.8383376306551654 0.9134573154135032\n",
      "smd 0.30188674394803194 0.16956732553478018 0.7516848672977844\n",
      "smd 0.6089644011523921 0.5694267902523401 0.9090340324382645\n",
      "smd 0.14173227007254122 0.3253086987807626 0.6642701194911487\n",
      "smd 0.4846415900872509 0.5021608655025646 0.9018584228352674\n",
      "smd 0.4821428179049775 0.521558181469944 0.6619355904445448\n"
     ]
    }
   ],
   "source": [
    "for loader in [load_smd]:\n",
    "    datasets = loader(8, 4)\n",
    "    x_trains, x_tests, y_tests = datasets['x_train'], datasets['x_test'], datasets['y_test']\n",
    "    \n",
    "    for i in tqdm(range(len(x_trains))):\n",
    "        tf.keras.backend.clear_session()\n",
    "\n",
    "        X_train = x_trains[i]\n",
    "        X_test = x_tests[i]\n",
    "        \n",
    "        gan = get_gan(X_train)\n",
    "        dataset = tf.data.Dataset.from_tensor_slices(X_train)\n",
    "        dataset = dataset.batch(128, drop_remainder=True).prefetch(1)\n",
    "\n",
    "        train_gan(gan, dataset, 128, X_train.shape[1], X_train.shape[2])\n",
    "        \n",
    "        X_test_rec = gan.layers[0].predict(X_test)\n",
    "        scores = evaluate(X_test, X_test_rec, y_tests[i], is_reconstructed=True)\n",
    "\n",
    "        total_scores['dataset'].append(loader.__name__.replace('load_', ''))\n",
    "        total_scores['f1'].append(np.max(scores['f1']))\n",
    "        total_scores['pr_auc'].append(scores['pr_auc'])\n",
    "        total_scores['roc_auc'].append(scores['roc_auc'])\n",
    "        print(loader.__name__.replace('load_', ''), np.max(scores['f1']), scores['pr_auc'], scores['roc_auc'])   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "23474773-bac1-4145-ac75-c1a686bca5c6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>f1</th>\n",
       "      <th>pr_auc</th>\n",
       "      <th>roc_auc</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dataset</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>smd</th>\n",
       "      <td>0.429715</td>\n",
       "      <td>0.418847</td>\n",
       "      <td>0.704664</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               f1    pr_auc   roc_auc\n",
       "dataset                              \n",
       "smd      0.429715  0.418847  0.704664"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "smd_results = pd.DataFrame(total_scores)\n",
    "smd_results.groupby('dataset').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "81ff42a3-d072-4fa5-973f-fc4747d315de",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "1acf69c6-5136-40b1-a62e-0057ab6e8e79",
   "metadata": {},
   "source": [
    "### ECG"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "110fcd09-f039-43a6-a27a-3b05f17e24fd",
   "metadata": {},
   "outputs": [],
   "source": [
    "total_scores = {'dataset': [], 'f1': [], 'pr_auc': [], 'roc_auc': []}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "33e7998d-10e7-48c4-b2a4-c5e841abc77c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a0d8f14d95b247598cfd63f8ebb2eff4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/9 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "badd52bc764d4393b9263511b0090848",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/9 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "D1 0.31364558118309294 0.2890681241493945 0.5080246213561493\n",
      "D2 0.22406637008316144 0.20242131471172264 0.34075421927235733\n",
      "D3 0.12087907210483784 0.049795927353707886 0.4461891356888046\n",
      "D4 0.24066386044318078 0.14990164599380668 0.5596859520306788\n",
      "D5 0.18848162848607283 0.12600726310766402 0.466154375602586\n",
      "D6 0.1875636928046768 0.08323584640605067 0.46833029881943244\n",
      "D7 0.07740773217129189 0.030190114458590354 0.6988161167403244\n",
      "D8 0.22200195981577328 0.09847685127039103 0.5560941163774945\n",
      "D9 0.36328121982822664 0.1870173081917467 0.44047309862533557\n"
     ]
    }
   ],
   "source": [
    "for loader in [load_ecg]:\n",
    "    datasets = loader(4, 2)\n",
    "    x_trains, x_tests, y_tests = datasets['x_train'], datasets['x_test'], datasets['y_test']\n",
    "    \n",
    "    for i in tqdm(range(len(x_trains))):\n",
    "        tf.keras.backend.clear_session()\n",
    "\n",
    "        X_train = x_trains[i]\n",
    "        X_test = x_tests[i]\n",
    "        \n",
    "        gan = get_gan(X_train)\n",
    "        dataset = tf.data.Dataset.from_tensor_slices(X_train)\n",
    "        dataset = dataset.batch(128, drop_remainder=True).prefetch(1)\n",
    "\n",
    "        train_gan(gan, dataset, 128, X_train.shape[1], X_train.shape[2])\n",
    "        \n",
    "        X_test_rec = gan.layers[0].predict(X_test)\n",
    "        scores = evaluate(X_test, X_test_rec, y_tests[i], is_reconstructed=True)\n",
    "        \n",
    "        total_scores['dataset'].append(f'D{i+1}')\n",
    "        total_scores['f1'].append(np.max(scores['f1']))\n",
    "        total_scores['pr_auc'].append(scores['pr_auc'])\n",
    "        total_scores['roc_auc'].append(scores['roc_auc'])\n",
    "        print(f'D{i+1}', np.max(scores['f1']), scores['pr_auc'], scores['roc_auc'])  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "dd066703-fab4-4dc0-8e8f-57fc2a9a7ff4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>f1</th>\n",
       "      <th>pr_auc</th>\n",
       "      <th>roc_auc</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dataset</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>D1</th>\n",
       "      <td>0.313646</td>\n",
       "      <td>0.289068</td>\n",
       "      <td>0.508025</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>D2</th>\n",
       "      <td>0.224066</td>\n",
       "      <td>0.202421</td>\n",
       "      <td>0.340754</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>D3</th>\n",
       "      <td>0.120879</td>\n",
       "      <td>0.049796</td>\n",
       "      <td>0.446189</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>D4</th>\n",
       "      <td>0.240664</td>\n",
       "      <td>0.149902</td>\n",
       "      <td>0.559686</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>D5</th>\n",
       "      <td>0.188482</td>\n",
       "      <td>0.126007</td>\n",
       "      <td>0.466154</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>D6</th>\n",
       "      <td>0.187564</td>\n",
       "      <td>0.083236</td>\n",
       "      <td>0.468330</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>D7</th>\n",
       "      <td>0.077408</td>\n",
       "      <td>0.030190</td>\n",
       "      <td>0.698816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>D8</th>\n",
       "      <td>0.222002</td>\n",
       "      <td>0.098477</td>\n",
       "      <td>0.556094</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>D9</th>\n",
       "      <td>0.363281</td>\n",
       "      <td>0.187017</td>\n",
       "      <td>0.440473</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               f1    pr_auc   roc_auc\n",
       "dataset                              \n",
       "D1       0.313646  0.289068  0.508025\n",
       "D2       0.224066  0.202421  0.340754\n",
       "D3       0.120879  0.049796  0.446189\n",
       "D4       0.240664  0.149902  0.559686\n",
       "D5       0.188482  0.126007  0.466154\n",
       "D6       0.187564  0.083236  0.468330\n",
       "D7       0.077408  0.030190  0.698816\n",
       "D8       0.222002  0.098477  0.556094\n",
       "D9       0.363281  0.187017  0.440473"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ecg_results = pd.DataFrame(total_scores)\n",
    "ecg_results.groupby('dataset').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "27c07cfe-19c3-4bcf-8c5c-19a70519d4d2",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "32b55486-3ced-4ecd-8b4b-3e7faf10abe4",
   "metadata": {},
   "source": [
    "### Power Demand"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "83b22ae9-1df8-4d9c-8db1-e4f719476df5",
   "metadata": {},
   "outputs": [],
   "source": [
    "total_scores = {'dataset': [], 'f1': [], 'pr_auc': [], 'roc_auc': []}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "5ee504a8-18a9-45ec-ae6a-387138d55736",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "316ed9854f9641f8800fa309643524c5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "power_demand 0.4523809085600946 0.37423619621403925 0.7438679617011319\n"
     ]
    }
   ],
   "source": [
    "for loader in [load_power_demand]:\n",
    "    datasets = loader(16, 8)\n",
    "    x_trains, x_tests, y_tests = datasets['x_train'], datasets['x_test'], datasets['y_test']\n",
    "    \n",
    "    for i in tqdm(range(len(x_trains))):\n",
    "        tf.keras.backend.clear_session()\n",
    "\n",
    "        X_train = x_trains[i]\n",
    "        X_test = x_tests[i]\n",
    "        \n",
    "        gan = get_gan(X_train)\n",
    "        dataset = tf.data.Dataset.from_tensor_slices(X_train)\n",
    "        dataset = dataset.batch(128, drop_remainder=True).prefetch(1)\n",
    "\n",
    "        train_gan(gan, dataset, 128, X_train.shape[1], X_train.shape[2])\n",
    "        \n",
    "        X_test_rec = gan.layers[0].predict(X_test)\n",
    "        scores = evaluate(X_test, X_test_rec, y_tests[i], is_reconstructed=True)\n",
    "\n",
    "        total_scores['dataset'].append(loader.__name__.replace('load_', ''))\n",
    "        total_scores['f1'].append(np.max(scores['f1']))\n",
    "        total_scores['pr_auc'].append(scores['pr_auc'])\n",
    "        total_scores['roc_auc'].append(scores['roc_auc'])\n",
    "        print(loader.__name__.replace('load_', ''), np.max(scores['f1']), scores['pr_auc'], scores['roc_auc']) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "96a9ddc1-cd03-4d47-8586-d9ac8ac25dea",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>f1</th>\n",
       "      <th>pr_auc</th>\n",
       "      <th>roc_auc</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dataset</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>power_demand</th>\n",
       "      <td>0.452381</td>\n",
       "      <td>0.374236</td>\n",
       "      <td>0.743868</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    f1    pr_auc   roc_auc\n",
       "dataset                                   \n",
       "power_demand  0.452381  0.374236  0.743868"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "power_results = pd.DataFrame(total_scores)\n",
    "power_results.groupby('dataset').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3aa398a5-c76e-4e5b-8b9b-4c255c66f756",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "21a2dd28-adbe-432c-9dfc-5fb478728d1b",
   "metadata": {},
   "source": [
    "### 2D Gesture"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "335b8cb2-b416-4850-8888-68af70886b73",
   "metadata": {},
   "outputs": [],
   "source": [
    "total_scores = {'dataset': [], 'f1': [], 'pr_auc': [], 'roc_auc': []}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "b1a951c6-fdc5-4b0c-be0b-244e57b4c024",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d269deca10504f86b732d0ea3f672493",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "gesture 0.4444444078974057 0.3424240523177565 0.5506097118696452\n"
     ]
    }
   ],
   "source": [
    "for loader in [load_gesture]:\n",
    "    datasets = loader(4, 2)\n",
    "    x_trains, x_tests, y_tests = datasets['x_train'], datasets['x_test'], datasets['y_test']\n",
    "    \n",
    "    for i in tqdm(range(len(x_trains))):\n",
    "        tf.keras.backend.clear_session()\n",
    "\n",
    "        X_train = x_trains[i]\n",
    "        X_test = x_tests[i]\n",
    "        \n",
    "        gan = get_gan(X_train)\n",
    "        dataset = tf.data.Dataset.from_tensor_slices(X_train)\n",
    "        dataset = dataset.batch(128, drop_remainder=True).prefetch(1)\n",
    "\n",
    "        train_gan(gan, dataset, 128, X_train.shape[1], X_train.shape[2])\n",
    "        \n",
    "        X_test_rec = gan.layers[0].predict(X_test)\n",
    "        scores = evaluate(X_test, X_test_rec, y_tests[i], is_reconstructed=True)\n",
    "\n",
    "\n",
    "        total_scores['dataset'].append(loader.__name__.replace('load_', ''))\n",
    "        total_scores['f1'].append(np.max(scores['f1']))\n",
    "        total_scores['pr_auc'].append(scores['pr_auc'])\n",
    "        total_scores['roc_auc'].append(scores['roc_auc'])\n",
    "        print(loader.__name__.replace('load_', ''), np.max(scores['f1']), scores['pr_auc'], scores['roc_auc'])  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "689bf1c8-2e34-4d9a-8a64-e17cd0b7df40",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>f1</th>\n",
       "      <th>pr_auc</th>\n",
       "      <th>roc_auc</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dataset</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>gesture</th>\n",
       "      <td>0.444444</td>\n",
       "      <td>0.342424</td>\n",
       "      <td>0.55061</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               f1    pr_auc  roc_auc\n",
       "dataset                             \n",
       "gesture  0.444444  0.342424  0.55061"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gesture_results = pd.DataFrame(total_scores)\n",
    "gesture_results.groupby('dataset').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "97fe0b29-c18a-46c6-8ce4-78e24d3dd470",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}